CCEP deploys Spot robots on production lines

Coca-Cola Europacific Partners has deployed Boston Dynamics Spot robots at its Wakefield and East Kilbride manufacturing sites to complete automated condition-monitoring rounds.

The mobile systems will inspect almost 600 points across seven production lines, collecting vibration, ultrasound, thermographic, and visual information while the plants remain in operation.

Maintenance teams can use the data to identify developing mechanical faults, compressed-air or gas leaks, heat anomalies, and other signs of deteriorating equipment before they produce an unplanned stoppage or quality problem.

Spot moves through industrial environments on articulated legs rather than wheels, allowing it to negotiate stairs, obstacles, narrow routes, and uneven floor conditions that can limit fixed sensors or conventional automated vehicles.

Repetitive inspection routes can be completed from comparable positions at regular intervals. Engineers are then able to concentrate on analysing deviations, planning repairs, and examining recurring faults rather than collecting every routine reading manually.

Thermal imaging can reveal overheating bearings, motors, gearboxes, and electrical connections, while ultrasound can locate compressed-air leaks that are difficult to hear in a noisy production hall. Vibration data supports the diagnosis of imbalance, looseness, misalignment, and bearing deterioration.

Condition data reshapes maintenance planning

Many factories combine planned servicing, operator observations, and reactive repairs. Fixed intervals provide structure, yet they may be too frequent for healthy equipment and too slow when a defect begins developing between inspections.

Condition monitoring changes the trigger by identifying movement away from a normal operating pattern. Labour, parts, and access can then be organised before the component fails, reducing the disruption associated with emergency intervention.

Beverage plants contain closely connected conveyors, fillers, cappers, labellers, packers, pumps, compressors, refrigeration systems, and utilities. A small component can stop a much larger line when accumulation capacity is exhausted or a critical process cannot be bypassed.

Repeated measurements also improve diagnosis. One elevated temperature may reflect ambient conditions or a temporary operating state, whereas a sustained upward trend across several rounds provides stronger evidence of deterioration.

Compressed-air losses are a particularly valuable target because they can continue unnoticed while increasing electricity use. Ultrasound provides a more reliable method of locating leaks, and repeat inspections can confirm whether repairs have remained effective.

Carbon dioxide and other gases require close control for cost, safety, and product quality. Mobile sensing and imagery can support investigation, although the robot must remain within established procedures for hazardous areas and restricted zones.

Food-machinery investment is increasingly being assessed through labour reduction, sanitation, data, and uptime alongside capacity. The wider processing and packaging machinery market is moving towards connected equipment capable of supporting maintenance and operational decisions throughout its life.

Integration will determine the return

Automated rounds can create large volumes of information, but the readings need thresholds, asset identities, historical context, and a defined route into maintenance planning. An anomaly that remains within a separate robotics dashboard is unlikely to change factory performance.

Connection with a computerised maintenance-management system would allow alerts, inspections, repairs, and component histories to be viewed together. Engineers could then compare a developing fault with previous failures and determine whether the same asset is producing a recurring pattern.

Thresholds must also reflect changing production conditions. Line speed, product format, ambient temperature, cleaning, and equipment loading can all alter vibration or heat, so an alarm system calibrated against one operating state may generate excessive warnings in another.

Too many poorly configured alerts create noise and weaken confidence. Baselines need to be built across normal production, with priorities separating conditions requiring immediate action from those suitable for review during planned maintenance.

The robots themselves become operational-technology assets requiring route management, cyber security, battery charging, calibration, software support, and controlled access to plant networks.

Factory layouts also change continually as pallets, hoses, temporary barriers, cleaning operations, and maintenance work alter the route. Safe deployment depends on the robot’s perception systems and on clear procedures governing where autonomous rounds can operate.

Hygiene zoning introduces another consideration because mobile equipment passing between areas can carry contamination. Routes, wheels or feet, charging stations, and cleaning methods must be compatible with the site’s food-safety controls.

The most effective arrangement is likely to combine permanent sensors, machine-control information, operator knowledge, and mobile inspection. Critical assets may warrant continuous fixed monitoring, while Spot can cover a larger group of machines where permanent installation would be costly or inflexible.

Deployment at two factories gives CCEP an opportunity to compare performance across different layouts, utilities, and line configurations. Almost 600 inspection points provide enough scale to assess whether the approach can move beyond a limited trial.

Fewer unexpected failures, shorter investigations, lower energy losses, and more focused use of engineering labour will provide the meaningful measures. The robots’ movement attracts attention, but the operational return will come from the maintenance decisions made after each round.